1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2016 Intel Corporation. 3 4.. _testpmd_runtime: 5 6Testpmd Runtime Functions 7========================= 8 9Where the testpmd application is started in interactive mode, (``-i|--interactive``), 10it displays a prompt that can be used to start and stop forwarding, 11configure the application, display statistics (including the extended NIC 12statistics aka xstats) , set the Flow Director and other tasks:: 13 14 testpmd> 15 16The testpmd prompt has some, limited, readline support. 17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 18as well as access to the command history via the up-arrow. 19 20There is also support for tab completion. 21If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 22 23.. code-block:: console 24 25 testpmd> show port <TAB> 26 27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 31 ... 32 33 34.. note:: 35 36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes:: 37 38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 39 (pause_time) (send_xon) (port_id) 40 41In the real ``testpmd>`` prompt these commands should be on a single line. 42 43Help Functions 44-------------- 45 46The testpmd has on-line help for the functions that are available at runtime. 47These are divided into sections and can be accessed using help, help section or help all: 48 49.. code-block:: console 50 51 testpmd> help 52 53 help control : Start and stop forwarding. 54 help display : Displaying port, stats and config information. 55 help config : Configuration information. 56 help ports : Configuring ports. 57 help registers : Reading and setting port registers. 58 help filters : Filters configuration help. 59 help all : All of the above sections. 60 61 62Command File Functions 63---------------------- 64 65To facilitate loading large number of commands or to avoid cutting and pasting where not 66practical or possible testpmd supports alternative methods for executing commands. 67 68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd 69 will execute all CLI commands contained within the file immediately before 70 starting packet forwarding or entering interactive mode. 71 72.. code-block:: console 73 74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt 75 Interactive-mode selected 76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt 77 Configuring Port 0 (socket 0) 78 Port 0: 7C:FE:90:CB:74:CE 79 Configuring Port 1 (socket 0) 80 Port 1: 7C:FE:90:CB:74:CA 81 Checking link statuses... 82 Port 0 Link Up - speed 10000 Mbps - full-duplex 83 Port 1 Link Up - speed 10000 Mbps - full-duplex 84 Done 85 Flow rule #0 created 86 Flow rule #1 created 87 ... 88 ... 89 Flow rule #498 created 90 Flow rule #499 created 91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 92 testpmd> 93 94 95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME`` 96 command. 97 98.. code-block:: console 99 100 testpmd> load /home/ubuntu/flow-create-commands.txt 101 Flow rule #0 created 102 Flow rule #1 created 103 ... 104 ... 105 Flow rule #498 created 106 Flow rule #499 created 107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 108 testpmd> 109 110 111In all cases output from any included command will be displayed as standard output. 112Execution will continue until the end of the file is reached regardless of 113whether any errors occur. The end user must examine the output to determine if 114any failures occurred. 115 116 117Control Functions 118----------------- 119 120start 121~~~~~ 122 123Start packet forwarding with current configuration:: 124 125 testpmd> start 126 127start tx_first 128~~~~~~~~~~~~~~ 129 130Start packet forwarding with current configuration after sending specified number of bursts of packets:: 131 132 testpmd> start tx_first (""|burst_num) 133 134The default burst number is 1 when ``burst_num`` not presented. 135 136stop 137~~~~ 138 139Stop packet forwarding, and display accumulated statistics:: 140 141 testpmd> stop 142 143quit 144~~~~ 145 146Quit to prompt:: 147 148 testpmd> quit 149 150 151Display Functions 152----------------- 153 154The functions in the following sections are used to display information about the 155testpmd configuration or the NIC status. 156 157show port 158~~~~~~~~~ 159 160Display information for a given port or all ports:: 161 162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 163 164The available information categories are: 165 166* ``info``: General port information such as MAC address. 167 168* ``stats``: RX/TX statistics. 169 170* ``xstats``: RX/TX extended NIC statistics. 171 172* ``fdir``: Flow Director information and statistics. 173 174* ``stat_qmap``: Queue statistics mapping. 175 176* ``dcb_tc``: DCB information such as TC mapping. 177 178* ``cap``: Supported offload capabilities. 179 180For example: 181 182.. code-block:: console 183 184 testpmd> show port info 0 185 186 ********************* Infos for port 0 ********************* 187 188 MAC address: XX:XX:XX:XX:XX:XX 189 Connect to socket: 0 190 memory allocation on the socket: 0 191 Link status: up 192 Link speed: 40000 Mbps 193 Link duplex: full-duplex 194 Promiscuous mode: enabled 195 Allmulticast mode: disabled 196 Maximum number of MAC addresses: 64 197 Maximum number of MAC addresses of hash filtering: 0 198 VLAN offload: 199 strip on 200 filter on 201 qinq(extend) off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 220show port rss reta 221~~~~~~~~~~~~~~~~~~ 222 223Display the rss redirection table entry indicated by masks on port X:: 224 225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 226 227size is used to indicate the hardware supported reta size 228 229show port rss-hash 230~~~~~~~~~~~~~~~~~~ 231 232Display the RSS hash functions and RSS hash key of a port:: 233 234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 235 236clear port 237~~~~~~~~~~ 238 239Clear the port statistics for a given port or for all ports:: 240 241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 242 243For example:: 244 245 testpmd> clear port stats all 246 247show (rxq|txq) 248~~~~~~~~~~~~~~ 249 250Display information for a given port's RX/TX queue:: 251 252 testpmd> show (rxq|txq) info (port_id) (queue_id) 253 254show config 255~~~~~~~~~~~ 256 257Displays the configuration of the application. 258The configuration comes from the command-line, the runtime or the application defaults:: 259 260 testpmd> show config (rxtx|cores|fwd|txpkts) 261 262The available information categories are: 263 264* ``rxtx``: RX/TX configuration items. 265 266* ``cores``: List of forwarding cores. 267 268* ``fwd``: Packet forwarding configuration. 269 270* ``txpkts``: Packets to TX configuration. 271 272For example: 273 274.. code-block:: console 275 276 testpmd> show config rxtx 277 278 io packet forwarding - CRC stripping disabled - packets/burst=16 279 nb forwarding cores=2 - nb forwarding ports=1 280 RX queues=1 - RX desc=128 - RX free threshold=0 281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 282 TX queues=1 - TX desc=512 - TX free threshold=0 283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 284 TX RS bit threshold=0 - TXQ flags=0x0 285 286set fwd 287~~~~~~~ 288 289Set the packet forwarding mode:: 290 291 testpmd> set fwd (io|mac|macswap|flowgen| \ 292 rxonly|txonly|csum|icmpecho) (""|retry) 293 294``retry`` can be specified for forwarding engines except ``rx_only``. 295 296The available information categories are: 297 298* ``io``: Forwards packets "as-is" in I/O mode. 299 This is the fastest possible forwarding operation as it does not access packets data. 300 This is the default mode. 301 302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 306 307* ``macswap``: MAC swap forwarding mode. 308 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 309 310* ``flowgen``: Multi-flow generation mode. 311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 312 313* ``rxonly``: Receives packets but doesn't transmit them. 314 315* ``txonly``: Generates and transmits packets without receiving any. 316 317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 318 319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 320 321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 322 323* ``tm``: Traffic Management forwarding mode 324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for 325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is 326 available as an default option that can be enabled through CLI. The user can 327 also modify the default hierarchy or specify the new hierarchy through CLI for 328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``. 329 330Example:: 331 332 testpmd> set fwd rxonly 333 334 Set rxonly packet forwarding mode 335 336 337read rxd 338~~~~~~~~ 339 340Display an RX descriptor for a port RX queue:: 341 342 testpmd> read rxd (port_id) (queue_id) (rxd_id) 343 344For example:: 345 346 testpmd> read rxd 0 0 4 347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 348 349read txd 350~~~~~~~~ 351 352Display a TX descriptor for a port TX queue:: 353 354 testpmd> read txd (port_id) (queue_id) (txd_id) 355 356For example:: 357 358 testpmd> read txd 0 0 4 359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 360 361ddp get list 362~~~~~~~~~~~~ 363 364Get loaded dynamic device personalization (DDP) package info list:: 365 366 testpmd> ddp get list (port_id) 367 368ddp get info 369~~~~~~~~~~~~ 370 371Display information about dynamic device personalization (DDP) profile:: 372 373 testpmd> ddp get info (profile_path) 374 375show vf stats 376~~~~~~~~~~~~~ 377 378Display VF statistics:: 379 380 testpmd> show vf stats (port_id) (vf_id) 381 382clear vf stats 383~~~~~~~~~~~~~~ 384 385Reset VF statistics:: 386 387 testpmd> clear vf stats (port_id) (vf_id) 388 389show port pctype mapping 390~~~~~~~~~~~~~~~~~~~~~~~~ 391 392List all items from the pctype mapping table:: 393 394 testpmd> show port (port_id) pctype mapping 395 396 397Configuration Functions 398----------------------- 399 400The testpmd application can be configured from the runtime as well as from the command-line. 401 402This section details the available configuration functions that are available. 403 404.. note:: 405 406 Configuration changes only become active when forwarding is started/restarted. 407 408set default 409~~~~~~~~~~~ 410 411Reset forwarding to the default configuration:: 412 413 testpmd> set default 414 415set verbose 416~~~~~~~~~~~ 417 418Set the debug verbosity level:: 419 420 testpmd> set verbose (level) 421 422Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 423 424set log 425~~~~~~~ 426 427Set the log level for a log type:: 428 429 testpmd> set log global|(type) (level) 430 431Where: 432 433* ``type`` is the log name. 434 435* ``level`` is the log level. 436 437For example, to change the global log level:: 438 testpmd> set log global (level) 439 440Regexes can also be used for type. To change log level of user1, user2 and user3:: 441 testpmd> set log user[1-3] (level) 442 443set nbport 444~~~~~~~~~~ 445 446Set the number of ports used by the application: 447 448set nbport (num) 449 450This is equivalent to the ``--nb-ports`` command-line option. 451 452set nbcore 453~~~~~~~~~~ 454 455Set the number of cores used by the application:: 456 457 testpmd> set nbcore (num) 458 459This is equivalent to the ``--nb-cores`` command-line option. 460 461.. note:: 462 463 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 464 465set coremask 466~~~~~~~~~~~~ 467 468Set the forwarding cores hexadecimal mask:: 469 470 testpmd> set coremask (mask) 471 472This is equivalent to the ``--coremask`` command-line option. 473 474.. note:: 475 476 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 477 478set portmask 479~~~~~~~~~~~~ 480 481Set the forwarding ports hexadecimal mask:: 482 483 testpmd> set portmask (mask) 484 485This is equivalent to the ``--portmask`` command-line option. 486 487set burst 488~~~~~~~~~ 489 490Set number of packets per burst:: 491 492 testpmd> set burst (num) 493 494This is equivalent to the ``--burst command-line`` option. 495 496When retry is enabled, the transmit delay time and number of retries can also be set:: 497 498 testpmd> set burst tx delay (microseconds) retry (num) 499 500set txpkts 501~~~~~~~~~~ 502 503Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 504 505 testpmd> set txpkts (x[,y]*) 506 507Where x[,y]* represents a CSV list of values, without white space. 508 509set txsplit 510~~~~~~~~~~~ 511 512Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 513 514 testpmd> set txsplit (off|on|rand) 515 516Where: 517 518* ``off`` disable packet copy & split for CSUM mode. 519 520* ``on`` split outgoing packet into multiple segments. Size of each segment 521 and number of segments per packet is determined by ``set txpkts`` command 522 (see above). 523 524* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 525 526set corelist 527~~~~~~~~~~~~ 528 529Set the list of forwarding cores:: 530 531 testpmd> set corelist (x[,y]*) 532 533For example, to change the forwarding cores: 534 535.. code-block:: console 536 537 testpmd> set corelist 3,1 538 testpmd> show config fwd 539 540 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 541 Logical Core 3 (socket 0) forwards packets on 1 streams: 542 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 543 Logical Core 1 (socket 0) forwards packets on 1 streams: 544 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 545 546.. note:: 547 548 The cores are used in the same order as specified on the command line. 549 550set portlist 551~~~~~~~~~~~~ 552 553Set the list of forwarding ports:: 554 555 testpmd> set portlist (x[,y]*) 556 557For example, to change the port forwarding: 558 559.. code-block:: console 560 561 testpmd> set portlist 0,2,1,3 562 testpmd> show config fwd 563 564 io packet forwarding - ports=4 - cores=1 - streams=4 565 Logical Core 3 (socket 0) forwards packets on 4 streams: 566 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 567 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 568 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 569 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 570 571set tx loopback 572~~~~~~~~~~~~~~~ 573 574Enable/disable tx loopback:: 575 576 testpmd> set tx loopback (port_id) (on|off) 577 578set drop enable 579~~~~~~~~~~~~~~~ 580 581set drop enable bit for all queues:: 582 583 testpmd> set all queues drop (port_id) (on|off) 584 585set split drop enable (for VF) 586~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 587 588set split drop enable bit for VF from PF:: 589 590 testpmd> set vf split drop (port_id) (vf_id) (on|off) 591 592set mac antispoof (for VF) 593~~~~~~~~~~~~~~~~~~~~~~~~~~ 594 595Set mac antispoof for a VF from the PF:: 596 597 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 598 599set macsec offload 600~~~~~~~~~~~~~~~~~~ 601 602Enable/disable MACsec offload:: 603 604 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 605 testpmd> set macsec offload (port_id) off 606 607set macsec sc 608~~~~~~~~~~~~~ 609 610Configure MACsec secure connection (SC):: 611 612 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 613 614.. note:: 615 616 The pi argument is ignored for tx. 617 Check the NIC Datasheet for hardware limits. 618 619set macsec sa 620~~~~~~~~~~~~~ 621 622Configure MACsec secure association (SA):: 623 624 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 625 626.. note:: 627 628 The IDX value must be 0 or 1. 629 Check the NIC Datasheet for hardware limits. 630 631set broadcast mode (for VF) 632~~~~~~~~~~~~~~~~~~~~~~~~~~~ 633 634Set broadcast mode for a VF from the PF:: 635 636 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 637 638vlan set strip 639~~~~~~~~~~~~~~ 640 641Set the VLAN strip on a port:: 642 643 testpmd> vlan set strip (on|off) (port_id) 644 645vlan set stripq 646~~~~~~~~~~~~~~~ 647 648Set the VLAN strip for a queue on a port:: 649 650 testpmd> vlan set stripq (on|off) (port_id,queue_id) 651 652vlan set stripq (for VF) 653~~~~~~~~~~~~~~~~~~~~~~~~ 654 655Set VLAN strip for all queues in a pool for a VF from the PF:: 656 657 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 658 659vlan set insert (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set VLAN insert for a VF from the PF:: 663 664 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 665 666vlan set tag (for VF) 667~~~~~~~~~~~~~~~~~~~~~ 668 669Set VLAN tag for a VF from the PF:: 670 671 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 672 673vlan set antispoof (for VF) 674~~~~~~~~~~~~~~~~~~~~~~~~~~~ 675 676Set VLAN antispoof for a VF from the PF:: 677 678 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 679 680vlan set filter 681~~~~~~~~~~~~~~~ 682 683Set the VLAN filter on a port:: 684 685 testpmd> vlan set filter (on|off) (port_id) 686 687vlan set qinq 688~~~~~~~~~~~~~ 689 690Set the VLAN QinQ (extended queue in queue) on for a port:: 691 692 testpmd> vlan set qinq (on|off) (port_id) 693 694vlan set tpid 695~~~~~~~~~~~~~ 696 697Set the inner or outer VLAN TPID for packet filtering on a port:: 698 699 testpmd> vlan set (inner|outer) tpid (value) (port_id) 700 701.. note:: 702 703 TPID value must be a 16-bit number (value <= 65536). 704 705rx_vlan add 706~~~~~~~~~~~ 707 708Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 709 710 testpmd> rx_vlan add (vlan_id|all) (port_id) 711 712.. note:: 713 714 VLAN filter must be set on that port. VLAN ID < 4096. 715 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 716 in VFTA table. This is important if enabling all vlan_ids. 717 718rx_vlan rm 719~~~~~~~~~~ 720 721Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 722 723 testpmd> rx_vlan rm (vlan_id|all) (port_id) 724 725rx_vlan add (for VF) 726~~~~~~~~~~~~~~~~~~~~ 727 728Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 729 730 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 731 732rx_vlan rm (for VF) 733~~~~~~~~~~~~~~~~~~~ 734 735Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 736 737 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 738 739tunnel_filter add 740~~~~~~~~~~~~~~~~~ 741 742Add a tunnel filter on a port:: 743 744 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 745 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 746 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 747 748The available information categories are: 749 750* ``vxlan``: Set tunnel type as VXLAN. 751 752* ``nvgre``: Set tunnel type as NVGRE. 753 754* ``ipingre``: Set tunnel type as IP-in-GRE. 755 756* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 757 758* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 759 760* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 761 762* ``imac``: Set filter type as Inner MAC. 763 764* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 765 766* ``oip``: Set filter type as Outer IP. 767 768* ``iip``: Set filter type as Inner IP. 769 770Example:: 771 772 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 773 192.168.2.2 0 ipingre oip 1 1 774 775 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 776 777tunnel_filter remove 778~~~~~~~~~~~~~~~~~~~~ 779 780Remove a tunnel filter on a port:: 781 782 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 783 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 784 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 785 786rx_vxlan_port add 787~~~~~~~~~~~~~~~~~ 788 789Add an UDP port for VXLAN packet filter on a port:: 790 791 testpmd> rx_vxlan_port add (udp_port) (port_id) 792 793rx_vxlan_port remove 794~~~~~~~~~~~~~~~~~~~~ 795 796Remove an UDP port for VXLAN packet filter on a port:: 797 798 testpmd> rx_vxlan_port rm (udp_port) (port_id) 799 800tx_vlan set 801~~~~~~~~~~~ 802 803Set hardware insertion of VLAN IDs in packets sent on a port:: 804 805 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 806 807For example, set a single VLAN ID (5) insertion on port 0:: 808 809 tx_vlan set 0 5 810 811Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 812 813 tx_vlan set 1 2 3 814 815 816tx_vlan set pvid 817~~~~~~~~~~~~~~~~ 818 819Set port based hardware insertion of VLAN ID in packets sent on a port:: 820 821 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 822 823tx_vlan reset 824~~~~~~~~~~~~~ 825 826Disable hardware insertion of a VLAN header in packets sent on a port:: 827 828 testpmd> tx_vlan reset (port_id) 829 830csum set 831~~~~~~~~ 832 833Select hardware or software calculation of the checksum when 834transmitting a packet using the ``csum`` forwarding engine:: 835 836 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 837 838Where: 839 840* ``ip|udp|tcp|sctp`` always relate to the inner layer. 841 842* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 843 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 844 supported). See also the ``csum parse-tunnel`` command. 845 846.. note:: 847 848 Check the NIC Datasheet for hardware limits. 849 850RSS queue region 851~~~~~~~~~~~~~~~~ 852 853Set RSS queue region span on a port:: 854 855 testpmd> set port (port_id) queue-region region_id (value) \ 856 queue_start_index (value) queue_num (value) 857 858Set flowtype mapping on a RSS queue region on a port:: 859 860 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 861 862where: 863 864* For the flowtype(pctype) of packet,the specific index for each type has 865 been defined in file i40e_type.h as enum i40e_filter_pctype. 866 867Set user priority mapping on a RSS queue region on a port:: 868 869 testpmd> set port (port_id) queue-region UP (value) region_id (value) 870 871Flush all queue region related configuration on a port:: 872 873 testpmd> set port (port_id) queue-region flush (on|off) 874 875where: 876 877* "on"is just an enable function which server for other configuration, 878 it is for all configuration about queue region from up layer, 879 at first will only keep in DPDK softwarestored in driver, 880 only after "flush on", it commit all configuration to HW. 881 "off" is just clean all configuration about queue region just now, 882 and restore all to DPDK i40e driver default config when start up. 883 884Show all queue region related configuration info on a port:: 885 886 testpmd> show port (port_id) queue-region 887 888.. note:: 889 890 Queue region only support on PF by now, so these command is 891 only for configuration of queue region on PF port. 892 893csum parse-tunnel 894~~~~~~~~~~~~~~~~~ 895 896Define how tunneled packets should be handled by the csum forward 897engine:: 898 899 testpmd> csum parse-tunnel (on|off) (tx_port_id) 900 901If enabled, the csum forward engine will try to recognize supported 902tunnel headers (vxlan, gre, ipip). 903 904If disabled, treat tunnel packets as non-tunneled packets (a inner 905header is handled as a packet payload). 906 907.. note:: 908 909 The port argument is the TX port like in the ``csum set`` command. 910 911Example: 912 913Consider a packet in packet like the following:: 914 915 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 916 917* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 918 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 919 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 920 921* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 922 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 923 924csum show 925~~~~~~~~~ 926 927Display tx checksum offload configuration:: 928 929 testpmd> csum show (port_id) 930 931tso set 932~~~~~~~ 933 934Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 935 936 testpmd> tso set (segsize) (port_id) 937 938.. note:: 939 940 Check the NIC datasheet for hardware limits. 941 942tso show 943~~~~~~~~ 944 945Display the status of TCP Segmentation Offload:: 946 947 testpmd> tso show (port_id) 948 949set port - gro 950~~~~~~~~~~~~~~ 951 952Enable or disable GRO in ``csum`` forwarding engine:: 953 954 testpmd> set port <port_id> gro on|off 955 956If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 957packets received from the given port. 958 959If disabled, packets received from the given port won't be performed 960GRO. By default, GRO is disabled for all ports. 961 962.. note:: 963 964 When enable GRO for a port, TCP/IPv4 packets received from the port 965 will be performed GRO. After GRO, all merged packets have bad 966 checksums, since the GRO library doesn't re-calculate checksums for 967 the merged packets. Therefore, if users want the merged packets to 968 have correct checksums, please select HW IP checksum calculation and 969 HW TCP checksum calculation for the port which the merged packets are 970 transmitted to. 971 972show port - gro 973~~~~~~~~~~~~~~~ 974 975Display GRO configuration for a given port:: 976 977 testpmd> show port <port_id> gro 978 979set gro flush 980~~~~~~~~~~~~~ 981 982Set the cycle to flush the GROed packets from reassembly tables:: 983 984 testpmd> set gro flush <cycles> 985 986When enable GRO, the csum forwarding engine performs GRO on received 987packets, and the GROed packets are stored in reassembly tables. Users 988can use this command to determine when the GROed packets are flushed 989from the reassembly tables. 990 991The ``cycles`` is measured in GRO operation times. The csum forwarding 992engine flushes the GROed packets from the tables every ``cycles`` GRO 993operations. 994 995By default, the value of ``cycles`` is 1, which means flush GROed packets 996from the reassembly tables as soon as one GRO operation finishes. The value 997of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 998 999Please note that the large value of ``cycles`` may cause the poor TCP/IP 1000stack performance. Because the GROed packets are delayed to arrive the 1001stack, thus causing more duplicated ACKs and TCP retransmissions. 1002 1003set port - gso 1004~~~~~~~~~~~~~~ 1005 1006Toggle per-port GSO support in ``csum`` forwarding engine:: 1007 1008 testpmd> set port <port_id> gso on|off 1009 1010If enabled, the csum forwarding engine will perform GSO on supported IPv4 1011packets, transmitted on the given port. 1012 1013If disabled, packets transmitted on the given port will not undergo GSO. 1014By default, GSO is disabled for all ports. 1015 1016.. note:: 1017 1018 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1019 port undergo GSO. Afterwards, the segmented packets are represented by 1020 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1021 of checksums for GSO'd segments in SW. As a result, if users want correct 1022 checksums in GSO segments, they should enable HW checksum calculation for 1023 GSO-enabled ports. 1024 1025 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1026 by setting the following options in the csum forwarding engine: 1027 1028 testpmd> csum set outer_ip hw <port_id> 1029 1030 testpmd> csum set ip hw <port_id> 1031 1032 testpmd> csum set tcp hw <port_id> 1033 1034set gso segsz 1035~~~~~~~~~~~~~ 1036 1037Set the maximum GSO segment size (measured in bytes), which includes the 1038packet header and the packet payload for GSO-enabled ports (global):: 1039 1040 testpmd> set gso segsz <length> 1041 1042show port - gso 1043~~~~~~~~~~~~~~~ 1044 1045Display the status of Generic Segmentation Offload for a given port:: 1046 1047 testpmd> show port <port_id> gso 1048 1049mac_addr add 1050~~~~~~~~~~~~ 1051 1052Add an alternative MAC address to a port:: 1053 1054 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1055 1056mac_addr remove 1057~~~~~~~~~~~~~~~ 1058 1059Remove a MAC address from a port:: 1060 1061 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1062 1063mac_addr add (for VF) 1064~~~~~~~~~~~~~~~~~~~~~ 1065 1066Add an alternative MAC address for a VF to a port:: 1067 1068 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1069 1070mac_addr set 1071~~~~~~~~~~~~ 1072 1073Set the default MAC address for a port:: 1074 1075 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1076 1077mac_addr set (for VF) 1078~~~~~~~~~~~~~~~~~~~~~ 1079 1080Set the MAC address for a VF from the PF:: 1081 1082 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1083 1084set eth-peer 1085~~~~~~~~~~~~ 1086 1087Set the forwarding peer address for certain port:: 1088 1089 testpmd> set eth-peer (port_id) (perr_addr) 1090 1091This is equivalent to the ``--eth-peer`` command-line option. 1092 1093set port-uta 1094~~~~~~~~~~~~ 1095 1096Set the unicast hash filter(s) on/off for a port:: 1097 1098 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1099 1100set promisc 1101~~~~~~~~~~~ 1102 1103Set the promiscuous mode on for a port or for all ports. 1104In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1105 1106 testpmd> set promisc (port_id|all) (on|off) 1107 1108set allmulti 1109~~~~~~~~~~~~ 1110 1111Set the allmulti mode for a port or for all ports:: 1112 1113 testpmd> set allmulti (port_id|all) (on|off) 1114 1115Same as the ifconfig (8) option. Controls how multicast packets are handled. 1116 1117set promisc (for VF) 1118~~~~~~~~~~~~~~~~~~~~ 1119 1120Set the unicast promiscuous mode for a VF from PF. 1121It's supported by Intel i40e NICs now. 1122In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1123 1124 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1125 1126set allmulticast (for VF) 1127~~~~~~~~~~~~~~~~~~~~~~~~~ 1128 1129Set the multicast promiscuous mode for a VF from PF. 1130It's supported by Intel i40e NICs now. 1131In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1132 1133 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1134 1135set tx max bandwidth (for VF) 1136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1137 1138Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1139 1140 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1141 1142set tc tx min bandwidth (for VF) 1143~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1144 1145Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1146 1147 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1148 1149set tc tx max bandwidth (for VF) 1150~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1151 1152Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1153 1154 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1155 1156set tc strict link priority mode 1157~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1158 1159Set some TCs' strict link priority mode on a physical port:: 1160 1161 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1162 1163set tc tx min bandwidth 1164~~~~~~~~~~~~~~~~~~~~~~~ 1165 1166Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1167 1168 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1169 1170set flow_ctrl rx 1171~~~~~~~~~~~~~~~~ 1172 1173Set the link flow control parameter on a port:: 1174 1175 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1176 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1177 autoneg (on|off) (port_id) 1178 1179Where: 1180 1181* ``high_water`` (integer): High threshold value to trigger XOFF. 1182 1183* ``low_water`` (integer): Low threshold value to trigger XON. 1184 1185* ``pause_time`` (integer): Pause quota in the Pause frame. 1186 1187* ``send_xon`` (0/1): Send XON frame. 1188 1189* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1190 1191* ``autoneg``: Change the auto-negotiation parameter. 1192 1193set pfc_ctrl rx 1194~~~~~~~~~~~~~~~ 1195 1196Set the priority flow control parameter on a port:: 1197 1198 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1199 (pause_time) (priority) (port_id) 1200 1201Where: 1202 1203* ``high_water`` (integer): High threshold value. 1204 1205* ``low_water`` (integer): Low threshold value. 1206 1207* ``pause_time`` (integer): Pause quota in the Pause frame. 1208 1209* ``priority`` (0-7): VLAN User Priority. 1210 1211set stat_qmap 1212~~~~~~~~~~~~~ 1213 1214Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1215 1216 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1217 1218For example, to set rx queue 2 on port 0 to mapping 5:: 1219 1220 testpmd>set stat_qmap rx 0 2 5 1221 1222set xstats-hide-zero 1223~~~~~~~~~~~~~~~~~~~~ 1224 1225Set the option to hide zero values for xstats display:: 1226 1227 testpmd> set xstats-hide-zero on|off 1228 1229.. note:: 1230 1231 By default, the zero values are displayed for xstats. 1232 1233set port - rx/tx (for VF) 1234~~~~~~~~~~~~~~~~~~~~~~~~~ 1235 1236Set VF receive/transmit from a port:: 1237 1238 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1239 1240set port - mac address filter (for VF) 1241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1242 1243Add/Remove unicast or multicast MAC addr filter for a VF:: 1244 1245 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1246 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1247 1248set port - rx mode(for VF) 1249~~~~~~~~~~~~~~~~~~~~~~~~~~ 1250 1251Set the VF receive mode of a port:: 1252 1253 testpmd> set port (port_id) vf (vf_id) \ 1254 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1255 1256The available receive modes are: 1257 1258* ``AUPE``: Accepts untagged VLAN. 1259 1260* ``ROPE``: Accepts unicast hash. 1261 1262* ``BAM``: Accepts broadcast packets. 1263 1264* ``MPE``: Accepts all multicast packets. 1265 1266set port - tx_rate (for Queue) 1267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1268 1269Set TX rate limitation for a queue on a port:: 1270 1271 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1272 1273set port - tx_rate (for VF) 1274~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1275 1276Set TX rate limitation for queues in VF on a port:: 1277 1278 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1279 1280set port - mirror rule 1281~~~~~~~~~~~~~~~~~~~~~~ 1282 1283Set pool or vlan type mirror rule for a port:: 1284 1285 testpmd> set port (port_id) mirror-rule (rule_id) \ 1286 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1287 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1288 1289Set link mirror rule for a port:: 1290 1291 testpmd> set port (port_id) mirror-rule (rule_id) \ 1292 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1293 1294For example to enable mirror traffic with vlan 0,1 to pool 0:: 1295 1296 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1297 1298reset port - mirror rule 1299~~~~~~~~~~~~~~~~~~~~~~~~ 1300 1301Reset a mirror rule for a port:: 1302 1303 testpmd> reset port (port_id) mirror-rule (rule_id) 1304 1305set flush_rx 1306~~~~~~~~~~~~ 1307 1308Set the flush on RX streams before forwarding. 1309The default is flush ``on``. 1310Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1311 1312 testpmd> set flush_rx off 1313 1314set bypass mode 1315~~~~~~~~~~~~~~~ 1316 1317Set the bypass mode for the lowest port on bypass enabled NIC:: 1318 1319 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1320 1321set bypass event 1322~~~~~~~~~~~~~~~~ 1323 1324Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1325 1326 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1327 mode (normal|bypass|isolate) (port_id) 1328 1329Where: 1330 1331* ``timeout``: Enable bypass after watchdog timeout. 1332 1333* ``os_on``: Enable bypass when OS/board is powered on. 1334 1335* ``os_off``: Enable bypass when OS/board is powered off. 1336 1337* ``power_on``: Enable bypass when power supply is turned on. 1338 1339* ``power_off``: Enable bypass when power supply is turned off. 1340 1341 1342set bypass timeout 1343~~~~~~~~~~~~~~~~~~ 1344 1345Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1346 1347 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1348 1349show bypass config 1350~~~~~~~~~~~~~~~~~~ 1351 1352Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1353 1354 testpmd> show bypass config (port_id) 1355 1356set link up 1357~~~~~~~~~~~ 1358 1359Set link up for a port:: 1360 1361 testpmd> set link-up port (port id) 1362 1363set link down 1364~~~~~~~~~~~~~ 1365 1366Set link down for a port:: 1367 1368 testpmd> set link-down port (port id) 1369 1370E-tag set 1371~~~~~~~~~ 1372 1373Enable E-tag insertion for a VF on a port:: 1374 1375 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1376 1377Disable E-tag insertion for a VF on a port:: 1378 1379 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1380 1381Enable/disable E-tag stripping on a port:: 1382 1383 testpmd> E-tag set stripping (on|off) port (port_id) 1384 1385Enable/disable E-tag based forwarding on a port:: 1386 1387 testpmd> E-tag set forwarding (on|off) port (port_id) 1388 1389Add an E-tag forwarding filter on a port:: 1390 1391 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1392 1393Delete an E-tag forwarding filter on a port:: 1394 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1395 1396ddp add 1397~~~~~~~ 1398 1399Load a dynamic device personalization (DDP) profile and store backup profile:: 1400 1401 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1402 1403ddp del 1404~~~~~~~ 1405 1406Delete a dynamic device personalization profile and restore backup profile:: 1407 1408 testpmd> ddp del (port_id) (backup_profile_path) 1409 1410ptype mapping 1411~~~~~~~~~~~~~ 1412 1413List all items from the ptype mapping table:: 1414 1415 testpmd> ptype mapping get (port_id) (valid_only) 1416 1417Where: 1418 1419* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1420 1421Replace a specific or a group of software defined ptype with a new one:: 1422 1423 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1424 1425where: 1426 1427* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1428 1429* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1430 1431* ``pkt_type``: The new software ptype to replace the old ones. 1432 1433Update hardware defined ptype to software defined packet type mapping table:: 1434 1435 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1436 1437where: 1438 1439* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1440 1441* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1442 1443Reset ptype mapping table:: 1444 1445 testpmd> ptype mapping reset (port_id) 1446 1447Port Functions 1448-------------- 1449 1450The following sections show functions for configuring ports. 1451 1452.. note:: 1453 1454 Port configuration changes only become active when forwarding is started/restarted. 1455 1456port attach 1457~~~~~~~~~~~ 1458 1459Attach a port specified by pci address or virtual device args:: 1460 1461 testpmd> port attach (identifier) 1462 1463To attach a new pci device, the device should be recognized by kernel first. 1464Then it should be moved under DPDK management. 1465Finally the port can be attached to testpmd. 1466 1467For example, to move a pci device using ixgbe under DPDK management: 1468 1469.. code-block:: console 1470 1471 # Check the status of the available devices. 1472 ./usertools/dpdk-devbind.py --status 1473 1474 Network devices using DPDK-compatible driver 1475 ============================================ 1476 <none> 1477 1478 Network devices using kernel driver 1479 =================================== 1480 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1481 1482 1483 # Bind the device to igb_uio. 1484 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1485 1486 1487 # Recheck the status of the devices. 1488 ./usertools/dpdk-devbind.py --status 1489 Network devices using DPDK-compatible driver 1490 ============================================ 1491 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1492 1493To attach a port created by virtual device, above steps are not needed. 1494 1495For example, to attach a port whose pci address is 0000:0a:00.0. 1496 1497.. code-block:: console 1498 1499 testpmd> port attach 0000:0a:00.0 1500 Attaching a new port... 1501 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1502 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1503 EAL: PCI memory mapped at 0x7f83bfa00000 1504 EAL: PCI memory mapped at 0x7f83bfa80000 1505 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1506 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1507 Port 0 is attached. Now total ports is 1 1508 Done 1509 1510For example, to attach a port created by pcap PMD. 1511 1512.. code-block:: console 1513 1514 testpmd> port attach net_pcap0 1515 Attaching a new port... 1516 PMD: Initializing pmd_pcap for net_pcap0 1517 PMD: Creating pcap-backed ethdev on numa socket 0 1518 Port 0 is attached. Now total ports is 1 1519 Done 1520 1521In this case, identifier is ``net_pcap0``. 1522This identifier format is the same as ``--vdev`` format of DPDK applications. 1523 1524For example, to re-attach a bonded port which has been previously detached, 1525the mode and slave parameters must be given. 1526 1527.. code-block:: console 1528 1529 testpmd> port attach net_bond_0,mode=0,slave=1 1530 Attaching a new port... 1531 EAL: Initializing pmd_bond for net_bond_0 1532 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1533 Port 0 is attached. Now total ports is 1 1534 Done 1535 1536 1537port detach 1538~~~~~~~~~~~ 1539 1540Detach a specific port:: 1541 1542 testpmd> port detach (port_id) 1543 1544Before detaching a port, the port should be stopped and closed. 1545 1546For example, to detach a pci device port 0. 1547 1548.. code-block:: console 1549 1550 testpmd> port stop 0 1551 Stopping ports... 1552 Done 1553 testpmd> port close 0 1554 Closing ports... 1555 Done 1556 1557 testpmd> port detach 0 1558 Detaching a port... 1559 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1560 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1561 EAL: PCI memory unmapped at 0x7f83bfa00000 1562 EAL: PCI memory unmapped at 0x7f83bfa80000 1563 Done 1564 1565 1566For example, to detach a virtual device port 0. 1567 1568.. code-block:: console 1569 1570 testpmd> port stop 0 1571 Stopping ports... 1572 Done 1573 testpmd> port close 0 1574 Closing ports... 1575 Done 1576 1577 testpmd> port detach 0 1578 Detaching a port... 1579 PMD: Closing pcap ethdev on numa socket 0 1580 Port 'net_pcap0' is detached. Now total ports is 0 1581 Done 1582 1583To remove a pci device completely from the system, first detach the port from testpmd. 1584Then the device should be moved under kernel management. 1585Finally the device can be removed using kernel pci hotplug functionality. 1586 1587For example, to move a pci device under kernel management: 1588 1589.. code-block:: console 1590 1591 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1592 1593 ./usertools/dpdk-devbind.py --status 1594 1595 Network devices using DPDK-compatible driver 1596 ============================================ 1597 <none> 1598 1599 Network devices using kernel driver 1600 =================================== 1601 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1602 1603To remove a port created by a virtual device, above steps are not needed. 1604 1605port start 1606~~~~~~~~~~ 1607 1608Start all ports or a specific port:: 1609 1610 testpmd> port start (port_id|all) 1611 1612port stop 1613~~~~~~~~~ 1614 1615Stop all ports or a specific port:: 1616 1617 testpmd> port stop (port_id|all) 1618 1619port close 1620~~~~~~~~~~ 1621 1622Close all ports or a specific port:: 1623 1624 testpmd> port close (port_id|all) 1625 1626port config - queue ring size 1627~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1628 1629Configure a rx/tx queue ring size:: 1630 1631 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1632 1633Only take effect after command that (re-)start the port or command that setup specific queue. 1634 1635port start/stop queue 1636~~~~~~~~~~~~~~~~~~~~~ 1637 1638Start/stop a rx/tx queue on a specific port:: 1639 1640 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1641 1642port setup queue 1643~~~~~~~~~~~~~~~~~~~~~ 1644 1645Setup a rx/tx queue on a specific port:: 1646 1647 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1648 1649Only take effect when port is started. 1650 1651port config - speed 1652~~~~~~~~~~~~~~~~~~~ 1653 1654Set the speed and duplex mode for all ports or a specific port:: 1655 1656 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1657 duplex (half|full|auto) 1658 1659port config - queues/descriptors 1660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1661 1662Set number of queues/descriptors for rxq, txq, rxd and txd:: 1663 1664 testpmd> port config all (rxq|txq|rxd|txd) (value) 1665 1666This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1667 1668port config - max-pkt-len 1669~~~~~~~~~~~~~~~~~~~~~~~~~ 1670 1671Set the maximum packet length:: 1672 1673 testpmd> port config all max-pkt-len (value) 1674 1675This is equivalent to the ``--max-pkt-len`` command-line option. 1676 1677port config - CRC Strip 1678~~~~~~~~~~~~~~~~~~~~~~~ 1679 1680Set hardware CRC stripping on or off for all ports:: 1681 1682 testpmd> port config all crc-strip (on|off) 1683 1684CRC stripping is on by default. 1685 1686The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1687 1688port config - scatter 1689~~~~~~~~~~~~~~~~~~~~~~~ 1690 1691Set RX scatter mode on or off for all ports:: 1692 1693 testpmd> port config all scatter (on|off) 1694 1695RX scatter mode is off by default. 1696 1697The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1698 1699port config - RX Checksum 1700~~~~~~~~~~~~~~~~~~~~~~~~~ 1701 1702Set hardware RX checksum offload to on or off for all ports:: 1703 1704 testpmd> port config all rx-cksum (on|off) 1705 1706Checksum offload is off by default. 1707 1708The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1709 1710port config - VLAN 1711~~~~~~~~~~~~~~~~~~ 1712 1713Set hardware VLAN on or off for all ports:: 1714 1715 testpmd> port config all hw-vlan (on|off) 1716 1717Hardware VLAN is off by default. 1718 1719The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1720 1721port config - VLAN filter 1722~~~~~~~~~~~~~~~~~~~~~~~~~ 1723 1724Set hardware VLAN filter on or off for all ports:: 1725 1726 testpmd> port config all hw-vlan-filter (on|off) 1727 1728Hardware VLAN filter is off by default. 1729 1730The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1731 1732port config - VLAN strip 1733~~~~~~~~~~~~~~~~~~~~~~~~ 1734 1735Set hardware VLAN strip on or off for all ports:: 1736 1737 testpmd> port config all hw-vlan-strip (on|off) 1738 1739Hardware VLAN strip is off by default. 1740 1741The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1742 1743port config - VLAN extend 1744~~~~~~~~~~~~~~~~~~~~~~~~~ 1745 1746Set hardware VLAN extend on or off for all ports:: 1747 1748 testpmd> port config all hw-vlan-extend (on|off) 1749 1750Hardware VLAN extend is off by default. 1751 1752The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1753 1754port config - Drop Packets 1755~~~~~~~~~~~~~~~~~~~~~~~~~~ 1756 1757Set packet drop for packets with no descriptors on or off for all ports:: 1758 1759 testpmd> port config all drop-en (on|off) 1760 1761Packet dropping for packets with no descriptors is off by default. 1762 1763The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1764 1765port config - RSS 1766~~~~~~~~~~~~~~~~~ 1767 1768Set the RSS (Receive Side Scaling) mode on or off:: 1769 1770 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1771 1772RSS is on by default. 1773 1774The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1775The ``default`` option enables all supported RSS types reported by device info. 1776The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1777 1778port config - RSS Reta 1779~~~~~~~~~~~~~~~~~~~~~~ 1780 1781Set the RSS (Receive Side Scaling) redirection table:: 1782 1783 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1784 1785port config - DCB 1786~~~~~~~~~~~~~~~~~ 1787 1788Set the DCB mode for an individual port:: 1789 1790 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1791 1792The traffic class should be 4 or 8. 1793 1794port config - Burst 1795~~~~~~~~~~~~~~~~~~~ 1796 1797Set the number of packets per burst:: 1798 1799 testpmd> port config all burst (value) 1800 1801This is equivalent to the ``--burst`` command-line option. 1802 1803port config - Threshold 1804~~~~~~~~~~~~~~~~~~~~~~~ 1805 1806Set thresholds for TX/RX queues:: 1807 1808 testpmd> port config all (threshold) (value) 1809 1810Where the threshold type can be: 1811 1812* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1813 1814* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1815 1816* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1817 1818* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1819 1820* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1821 1822* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1823 1824* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1825 1826* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1827 1828* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1829 1830These threshold options are also available from the command-line. 1831 1832port config - E-tag 1833~~~~~~~~~~~~~~~~~~~ 1834 1835Set the value of ether-type for E-tag:: 1836 1837 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1838 1839Enable/disable the E-tag support:: 1840 1841 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1842 1843port config pctype mapping 1844~~~~~~~~~~~~~~~~~~~~~~~~~~ 1845 1846Reset pctype mapping table:: 1847 1848 testpmd> port config (port_id) pctype mapping reset 1849 1850Update hardware defined pctype to software defined flow type mapping table:: 1851 1852 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1853 1854where: 1855 1856* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1857 1858* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1859 1860port config input set 1861~~~~~~~~~~~~~~~~~~~~~ 1862 1863Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1864 testpmd> port config (port_id) pctype (pctype_id) \ 1865 (hash_inset|fdir_inset|fdir_flx_inset) \ 1866 (get|set|clear) field (field_idx) 1867 1868Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1869 testpmd> port config (port_id) pctype (pctype_id) \ 1870 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1871 1872where: 1873 1874* ``pctype_id``: hardware packet classification types. 1875* ``field_idx``: hardware field index. 1876 1877port config udp_tunnel_port 1878~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1879 1880Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 1881 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 1882 1883Link Bonding Functions 1884---------------------- 1885 1886The Link Bonding functions make it possible to dynamically create and 1887manage link bonding devices from within testpmd interactive prompt. 1888 1889create bonded device 1890~~~~~~~~~~~~~~~~~~~~ 1891 1892Create a new bonding device:: 1893 1894 testpmd> create bonded device (mode) (socket) 1895 1896For example, to create a bonded device in mode 1 on socket 0:: 1897 1898 testpmd> create bonded 1 0 1899 created new bonded device (port X) 1900 1901add bonding slave 1902~~~~~~~~~~~~~~~~~ 1903 1904Adds Ethernet device to a Link Bonding device:: 1905 1906 testpmd> add bonding slave (slave id) (port id) 1907 1908For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 1909 1910 testpmd> add bonding slave 6 10 1911 1912 1913remove bonding slave 1914~~~~~~~~~~~~~~~~~~~~ 1915 1916Removes an Ethernet slave device from a Link Bonding device:: 1917 1918 testpmd> remove bonding slave (slave id) (port id) 1919 1920For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 1921 1922 testpmd> remove bonding slave 6 10 1923 1924set bonding mode 1925~~~~~~~~~~~~~~~~ 1926 1927Set the Link Bonding mode of a Link Bonding device:: 1928 1929 testpmd> set bonding mode (value) (port id) 1930 1931For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 1932 1933 testpmd> set bonding mode 3 10 1934 1935set bonding primary 1936~~~~~~~~~~~~~~~~~~~ 1937 1938Set an Ethernet slave device as the primary device on a Link Bonding device:: 1939 1940 testpmd> set bonding primary (slave id) (port id) 1941 1942For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 1943 1944 testpmd> set bonding primary 6 10 1945 1946set bonding mac 1947~~~~~~~~~~~~~~~ 1948 1949Set the MAC address of a Link Bonding device:: 1950 1951 testpmd> set bonding mac (port id) (mac) 1952 1953For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 1954 1955 testpmd> set bonding mac 10 00:00:00:00:00:01 1956 1957set bonding xmit_balance_policy 1958~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1959 1960Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 1961 1962 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 1963 1964For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 1965 1966 testpmd> set bonding xmit_balance_policy 10 l34 1967 1968 1969set bonding mon_period 1970~~~~~~~~~~~~~~~~~~~~~~ 1971 1972Set the link status monitoring polling period in milliseconds for a bonding device. 1973 1974This adds support for PMD slave devices which do not support link status interrupts. 1975When the mon_period is set to a value greater than 0 then all PMD's which do not support 1976link status ISR will be queried every polling interval to check if their link status has changed:: 1977 1978 testpmd> set bonding mon_period (port_id) (value) 1979 1980For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 1981 1982 testpmd> set bonding mon_period 5 150 1983 1984 1985set bonding lacp dedicated_queue 1986~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1987 1988Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 1989when in mode 4 (link-aggregration-802.3ad):: 1990 1991 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 1992 1993 1994set bonding agg_mode 1995~~~~~~~~~~~~~~~~~~~~ 1996 1997Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 1998 1999 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2000 2001 2002show bonding config 2003~~~~~~~~~~~~~~~~~~~ 2004 2005Show the current configuration of a Link Bonding device:: 2006 2007 testpmd> show bonding config (port id) 2008 2009For example, 2010to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2011in balance mode with a transmission policy of layer 2+3:: 2012 2013 testpmd> show bonding config 9 2014 Bonding mode: 2 2015 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2016 Slaves (3): [1 3 4] 2017 Active Slaves (3): [1 3 4] 2018 Primary: [3] 2019 2020 2021Register Functions 2022------------------ 2023 2024The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2025This is mainly useful for debugging purposes. 2026Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2027and fields that can be accessed. 2028 2029read reg 2030~~~~~~~~ 2031 2032Display the value of a port register:: 2033 2034 testpmd> read reg (port_id) (address) 2035 2036For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2037 2038 testpmd> read reg 0 0xEE00 2039 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2040 2041read regfield 2042~~~~~~~~~~~~~ 2043 2044Display a port register bit field:: 2045 2046 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2047 2048For example, reading the lowest two bits from the register in the example above:: 2049 2050 testpmd> read regfield 0 0xEE00 0 1 2051 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2052 2053read regbit 2054~~~~~~~~~~~ 2055 2056Display a single port register bit:: 2057 2058 testpmd> read regbit (port_id) (address) (bit_x) 2059 2060For example, reading the lowest bit from the register in the example above:: 2061 2062 testpmd> read regbit 0 0xEE00 0 2063 port 0 PCI register at offset 0xEE00: bit 0=1 2064 2065write reg 2066~~~~~~~~~ 2067 2068Set the value of a port register:: 2069 2070 testpmd> write reg (port_id) (address) (value) 2071 2072For example, to clear a register:: 2073 2074 testpmd> write reg 0 0xEE00 0x0 2075 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2076 2077write regfield 2078~~~~~~~~~~~~~~ 2079 2080Set bit field of a port register:: 2081 2082 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2083 2084For example, writing to the register cleared in the example above:: 2085 2086 testpmd> write regfield 0 0xEE00 0 1 2 2087 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2088 2089write regbit 2090~~~~~~~~~~~~ 2091 2092Set single bit value of a port register:: 2093 2094 testpmd> write regbit (port_id) (address) (bit_x) (value) 2095 2096For example, to set the high bit in the register from the example above:: 2097 2098 testpmd> write regbit 0 0xEE00 31 1 2099 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2100 2101Traffic Metering and Policing 2102----------------------------- 2103 2104The following section shows functions for configuring traffic metering and 2105policing on the ethernet device through the use of generic ethdev API. 2106 2107show port traffic management capability 2108~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2109 2110Show traffic metering and policing capability of the port:: 2111 2112 testpmd> show port meter cap (port_id) 2113 2114add port meter profile (srTCM rfc2967) 2115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2116 2117Add meter profile (srTCM rfc2697) to the ethernet device:: 2118 2119 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2120 (cir) (cbs) (ebs) 2121 2122where: 2123 2124* ``profile_id``: ID for the meter profile. 2125* ``cir``: Committed Information Rate (CIR) (bytes/second). 2126* ``cbs``: Committed Burst Size (CBS) (bytes). 2127* ``ebs``: Excess Burst Size (EBS) (bytes). 2128 2129add port meter profile (trTCM rfc2968) 2130~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2131 2132Add meter profile (srTCM rfc2698) to the ethernet device:: 2133 2134 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2135 (cir) (pir) (cbs) (pbs) 2136 2137where: 2138 2139* ``profile_id``: ID for the meter profile. 2140* ``cir``: Committed information rate (bytes/second). 2141* ``pir``: Peak information rate (bytes/second). 2142* ``cbs``: Committed burst size (bytes). 2143* ``pbs``: Peak burst size (bytes). 2144 2145add port meter profile (trTCM rfc4115) 2146~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2147 2148Add meter profile (trTCM rfc4115) to the ethernet device:: 2149 2150 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2151 (cir) (eir) (cbs) (ebs) 2152 2153where: 2154 2155* ``profile_id``: ID for the meter profile. 2156* ``cir``: Committed information rate (bytes/second). 2157* ``eir``: Excess information rate (bytes/second). 2158* ``cbs``: Committed burst size (bytes). 2159* ``ebs``: Excess burst size (bytes). 2160 2161delete port meter profile 2162~~~~~~~~~~~~~~~~~~~~~~~~~ 2163 2164Delete meter profile from the ethernet device:: 2165 2166 testpmd> del port meter profile (port_id) (profile_id) 2167 2168create port meter 2169~~~~~~~~~~~~~~~~~ 2170 2171Create new meter object for the ethernet device:: 2172 2173 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2174 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2175 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2176 (dscp_tbl_entry63)] 2177 2178where: 2179 2180* ``mtr_id``: meter object ID. 2181* ``profile_id``: ID for the meter profile. 2182* ``meter_enable``: When this parameter has a non-zero value, the meter object 2183 gets enabled at the time of creation, otherwise remains disabled. 2184* ``g_action``: Policer action for the packet with green color. 2185* ``y_action``: Policer action for the packet with yellow color. 2186* ``r_action``: Policer action for the packet with red color. 2187* ``stats_mask``: Mask of statistics counter types to be enabled for the 2188 meter object. 2189* ``shared``: When this parameter has a non-zero value, the meter object is 2190 shared by multiple flows. Otherwise, meter object is used by single flow. 2191* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2192 input color for the current meter object is determined by the latest meter 2193 object in the same flow. Otherwise, the current meter object uses the 2194 *dscp_table* to determine the input color. 2195* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2196 color, 0 <= x <= 63. 2197 2198enable port meter 2199~~~~~~~~~~~~~~~~~ 2200 2201Enable meter for the ethernet device:: 2202 2203 testpmd> enable port meter (port_id) (mtr_id) 2204 2205disable port meter 2206~~~~~~~~~~~~~~~~~~ 2207 2208Disable meter for the ethernet device:: 2209 2210 testpmd> disable port meter (port_id) (mtr_id) 2211 2212delete port meter 2213~~~~~~~~~~~~~~~~~ 2214 2215Delete meter for the ethernet device:: 2216 2217 testpmd> del port meter (port_id) (mtr_id) 2218 2219Set port meter profile 2220~~~~~~~~~~~~~~~~~~~~~~ 2221 2222Set meter profile for the ethernet device:: 2223 2224 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2225 2226set port meter dscp table 2227~~~~~~~~~~~~~~~~~~~~~~~~~ 2228 2229Set meter dscp table for the ethernet device:: 2230 2231 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2232 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2233 2234set port meter policer action 2235~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2236 2237Set meter policer action for the ethernet device:: 2238 2239 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2240 (action0) [(action1) (action1)] 2241 2242where: 2243 2244* ``action_mask``: Bit mask indicating which policer actions need to be 2245 updated. One or more policer actions can be updated in a single function 2246 invocation. To update the policer action associated with color C, bit 2247 (1 << C) needs to be set in *action_mask* and element at position C 2248 in the *actions* array needs to be valid. 2249* ``actionx``: Policer action for the color x, 2250 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2251 2252set port meter stats mask 2253~~~~~~~~~~~~~~~~~~~~~~~~~ 2254 2255Set meter stats mask for the ethernet device:: 2256 2257 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2258 2259where: 2260 2261* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2262 2263show port meter stats 2264~~~~~~~~~~~~~~~~~~~~~ 2265 2266Show meter stats of the ethernet device:: 2267 2268 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2269 2270where: 2271 2272* ``clear``: Flag that indicates whether the statistics counters should 2273 be cleared (i.e. set to zero) immediately after they have been read or not. 2274 2275Traffic Management 2276------------------ 2277 2278The following section shows functions for configuring traffic management on 2279on the ethernet device through the use of generic TM API. 2280 2281show port traffic management capability 2282~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2283 2284Show traffic management capability of the port:: 2285 2286 testpmd> show port tm cap (port_id) 2287 2288show port traffic management capability (hierarchy level) 2289~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2290 2291Show traffic management hierarchy level capability of the port:: 2292 2293 testpmd> show port tm level cap (port_id) (level_id) 2294 2295show port traffic management capability (hierarchy node level) 2296~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2297 2298Show the traffic management hierarchy node capability of the port:: 2299 2300 testpmd> show port tm node cap (port_id) (node_id) 2301 2302show port traffic management hierarchy node type 2303~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2304 2305Show the port traffic management hierarchy node type:: 2306 2307 testpmd> show port tm node type (port_id) (node_id) 2308 2309show port traffic management hierarchy node stats 2310~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2311 2312Show the port traffic management hierarchy node statistics:: 2313 2314 testpmd> show port tm node stats (port_id) (node_id) (clear) 2315 2316where: 2317 2318* ``clear``: When this parameter has a non-zero value, the statistics counters 2319 are cleared (i.e. set to zero) immediately after they have been read, 2320 otherwise the statistics counters are left untouched. 2321 2322Add port traffic management private shaper profile 2323~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2324 2325Add the port traffic management private shaper profile:: 2326 2327 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2328 (tb_rate) (tb_size) (packet_length_adjust) 2329 2330where: 2331 2332* ``shaper_profile id``: Shaper profile ID for the new profile. 2333* ``tb_rate``: Token bucket rate (bytes per second). 2334* ``tb_size``: Token bucket size (bytes). 2335* ``packet_length_adjust``: The value (bytes) to be added to the length of 2336 each packet for the purpose of shaping. This parameter value can be used to 2337 correct the packet length with the framing overhead bytes that are consumed 2338 on the wire. 2339 2340Delete port traffic management private shaper profile 2341~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2342 2343Delete the port traffic management private shaper:: 2344 2345 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2346 2347where: 2348 2349* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2350 2351Add port traffic management shared shaper 2352~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2353 2354Create the port traffic management shared shaper:: 2355 2356 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2357 (shaper_profile_id) 2358 2359where: 2360 2361* ``shared_shaper_id``: Shared shaper ID to be created. 2362* ``shaper_profile id``: Shaper profile ID for shared shaper. 2363 2364Set port traffic management shared shaper 2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2366 2367Update the port traffic management shared shaper:: 2368 2369 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2370 (shaper_profile_id) 2371 2372where: 2373 2374* ``shared_shaper_id``: Shared shaper ID to be update. 2375* ``shaper_profile id``: Shaper profile ID for shared shaper. 2376 2377Delete port traffic management shared shaper 2378~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2379 2380Delete the port traffic management shared shaper:: 2381 2382 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2383 2384where: 2385 2386* ``shared_shaper_id``: Shared shaper ID to be deleted. 2387 2388Set port traffic management hiearchy node private shaper 2389~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2390 2391set the port traffic management hierarchy node private shaper:: 2392 2393 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2394 (shaper_profile_id) 2395 2396where: 2397 2398* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2399 hierarchy node. 2400 2401Add port traffic management WRED profile 2402~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2403 2404Create a new WRED profile:: 2405 2406 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2407 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2408 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2409 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2410 2411where: 2412 2413* ``wred_profile id``: Identifier for the newly create WRED profile 2414* ``color_g``: Packet color (green) 2415* ``min_th_g``: Minimum queue threshold for packet with green color 2416* ``max_th_g``: Minimum queue threshold for packet with green color 2417* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2418* ``wq_log2_g``: Negated log2 of queue weight (wq) 2419* ``color_y``: Packet color (yellow) 2420* ``min_th_y``: Minimum queue threshold for packet with yellow color 2421* ``max_th_y``: Minimum queue threshold for packet with yellow color 2422* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2423* ``wq_log2_y``: Negated log2 of queue weight (wq) 2424* ``color_r``: Packet color (red) 2425* ``min_th_r``: Minimum queue threshold for packet with yellow color 2426* ``max_th_r``: Minimum queue threshold for packet with yellow color 2427* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2428* ``wq_log2_r``: Negated log2 of queue weight (wq) 2429 2430Delete port traffic management WRED profile 2431~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2432 2433Delete the WRED profile:: 2434 2435 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2436 2437Add port traffic management hierarchy nonleaf node 2438~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2439 2440Add nonleaf node to port traffic management hiearchy:: 2441 2442 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2443 (priority) (weight) (level_id) (shaper_profile_id) \ 2444 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2445 [(shared_shaper_0) (shared_shaper_1) ...] \ 2446 2447where: 2448 2449* ``parent_node_id``: Node ID of the parent. 2450* ``priority``: Node priority (highest node priority is zero). This is used by 2451 the SP algorithm running on the parent node for scheduling this node. 2452* ``weight``: Node weight (lowest weight is one). The node weight is relative 2453 to the weight sum of all siblings that have the same priority. It is used by 2454 the WFQ algorithm running on the parent node for scheduling this node. 2455* ``level_id``: Hiearchy level of the node. 2456* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2457 the node. 2458* ``n_sp_priorities``: Number of strict priorities. 2459* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2460* ``n_shared_shapers``: Number of shared shapers. 2461* ``shared_shaper_id``: Shared shaper id. 2462 2463Add port traffic management hierarchy leaf node 2464~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2465 2466Add leaf node to port traffic management hiearchy:: 2467 2468 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2469 (priority) (weight) (level_id) (shaper_profile_id) \ 2470 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2471 [(shared_shaper_id) (shared_shaper_id) ...] \ 2472 2473where: 2474 2475* ``parent_node_id``: Node ID of the parent. 2476* ``priority``: Node priority (highest node priority is zero). This is used by 2477 the SP algorithm running on the parent node for scheduling this node. 2478* ``weight``: Node weight (lowest weight is one). The node weight is relative 2479 to the weight sum of all siblings that have the same priority. It is used by 2480 the WFQ algorithm running on the parent node for scheduling this node. 2481* ``level_id``: Hiearchy level of the node. 2482* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2483 the node. 2484* ``cman_mode``: Congestion management mode to be enabled for this node. 2485* ``wred_profile_id``: WRED profile id to be enabled for this node. 2486* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2487* ``n_shared_shapers``: Number of shared shapers. 2488* ``shared_shaper_id``: Shared shaper id. 2489 2490Delete port traffic management hierarchy node 2491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2492 2493Delete node from port traffic management hiearchy:: 2494 2495 testpmd> del port tm node (port_id) (node_id) 2496 2497Update port traffic management hierarchy parent node 2498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2499 2500Update port traffic management hierarchy parent node:: 2501 2502 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2503 (priority) (weight) 2504 2505This function can only be called after the hierarchy commit invocation. Its 2506success depends on the port support for this operation, as advertised through 2507the port capability set. This function is valid for all nodes of the traffic 2508management hierarchy except root node. 2509 2510Suspend port traffic management hierarchy node 2511~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2512 2513 testpmd> suspend port tm node (port_id) (node_id) 2514 2515Resume port traffic management hierarchy node 2516~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2517 2518 testpmd> resume port tm node (port_id) (node_id) 2519 2520Commit port traffic management hierarchy 2521~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2522 2523Commit the traffic management hierarchy on the port:: 2524 2525 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2526 2527where: 2528 2529* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2530 call failure. On the other hand, hierarchy is preserved when this parameter 2531 is equal to zero. 2532 2533Set port traffic management default hierarchy (tm forwarding mode) 2534~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2535 2536set the traffic management default hierarchy on the port:: 2537 2538 testpmd> set port tm hierarchy default (port_id) 2539 2540Filter Functions 2541---------------- 2542 2543This section details the available filter functions that are available. 2544 2545Note these functions interface the deprecated legacy filtering framework, 2546superseded by *rte_flow*. See `Flow rules management`_. 2547 2548ethertype_filter 2549~~~~~~~~~~~~~~~~~~~~ 2550 2551Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2552 2553 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2554 ethertype (ether_type) (drop|fwd) queue (queue_id) 2555 2556The available information parameters are: 2557 2558* ``port_id``: The port which the Ethertype filter assigned on. 2559 2560* ``mac_addr``: Compare destination mac address. 2561 2562* ``mac_ignr``: Ignore destination mac address match. 2563 2564* ``mac_address``: Destination mac address to match. 2565 2566* ``ether_type``: The EtherType value want to match, 2567 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2568 2569* ``queue_id``: The receive queue associated with this EtherType filter. 2570 It is meaningless when deleting or dropping. 2571 2572Example, to add/remove an ethertype filter rule:: 2573 2574 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2575 ethertype 0x0806 fwd queue 3 2576 2577 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2578 ethertype 0x0806 fwd queue 3 2579 25802tuple_filter 2581~~~~~~~~~~~~~~~~~ 2582 2583Add or delete a 2-tuple filter, 2584which identifies packets by specific protocol and destination TCP/UDP port 2585and forwards packets into one of the receive queues:: 2586 2587 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2588 protocol (protocol_value) mask (mask_value) \ 2589 tcp_flags (tcp_flags_value) priority (prio_value) \ 2590 queue (queue_id) 2591 2592The available information parameters are: 2593 2594* ``port_id``: The port which the 2-tuple filter assigned on. 2595 2596* ``dst_port_value``: Destination port in L4. 2597 2598* ``protocol_value``: IP L4 protocol. 2599 2600* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2601 2602* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2603 2604* ``prio_value``: Priority of this filter. 2605 2606* ``queue_id``: The receive queue associated with this 2-tuple filter. 2607 2608Example, to add/remove an 2tuple filter rule:: 2609 2610 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2611 tcp_flags 0x02 priority 3 queue 3 2612 2613 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2614 tcp_flags 0x02 priority 3 queue 3 2615 26165tuple_filter 2617~~~~~~~~~~~~~~~~~ 2618 2619Add or delete a 5-tuple filter, 2620which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2621and routes packets into one of the receive queues:: 2622 2623 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2624 (src_address) dst_port (dst_port_value) \ 2625 src_port (src_port_value) protocol (protocol_value) \ 2626 mask (mask_value) tcp_flags (tcp_flags_value) \ 2627 priority (prio_value) queue (queue_id) 2628 2629The available information parameters are: 2630 2631* ``port_id``: The port which the 5-tuple filter assigned on. 2632 2633* ``dst_address``: Destination IP address. 2634 2635* ``src_address``: Source IP address. 2636 2637* ``dst_port_value``: TCP/UDP destination port. 2638 2639* ``src_port_value``: TCP/UDP source port. 2640 2641* ``protocol_value``: L4 protocol. 2642 2643* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2644 2645* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2646 2647* ``prio_value``: The priority of this filter. 2648 2649* ``queue_id``: The receive queue associated with this 5-tuple filter. 2650 2651Example, to add/remove an 5tuple filter rule:: 2652 2653 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2654 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2655 flags 0x0 priority 3 queue 3 2656 2657 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2658 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2659 flags 0x0 priority 3 queue 3 2660 2661syn_filter 2662~~~~~~~~~~ 2663 2664Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2665 2666 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2667 2668The available information parameters are: 2669 2670* ``port_id``: The port which the SYN filter assigned on. 2671 2672* ``high``: This SYN filter has higher priority than other filters. 2673 2674* ``low``: This SYN filter has lower priority than other filters. 2675 2676* ``queue_id``: The receive queue associated with this SYN filter 2677 2678Example:: 2679 2680 testpmd> syn_filter 0 add priority high queue 3 2681 2682flex_filter 2683~~~~~~~~~~~ 2684 2685With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2686and routed into one of the receive queues:: 2687 2688 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2689 mask (mask_value) priority (prio_value) queue (queue_id) 2690 2691The available information parameters are: 2692 2693* ``port_id``: The port which the Flex filter is assigned on. 2694 2695* ``len_value``: Filter length in bytes, no greater than 128. 2696 2697* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2698 2699* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2700 2701* ``prio_value``: The priority of this filter. 2702 2703* ``queue_id``: The receive queue associated with this Flex filter. 2704 2705Example:: 2706 2707 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2708 mask 000C priority 3 queue 3 2709 2710 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2711 mask 000C priority 3 queue 3 2712 2713 2714.. _testpmd_flow_director: 2715 2716flow_director_filter 2717~~~~~~~~~~~~~~~~~~~~ 2718 2719The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2720 2721Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2722Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2723 2724* Perfect match filters. 2725 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2726 The masked fields are for IP flow. 2727 2728* Signature filters. 2729 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2730 2731* Perfect-mac-vlan match filters. 2732 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2733 The masked fields are for MAC VLAN flow. 2734 2735* Perfect-tunnel match filters. 2736 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2737 The masked fields are for tunnel flow. 2738 2739* Perfect-raw-flow-type match filters. 2740 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2741 The masked fields are specified by input sets. 2742 2743The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2744per flow type and the flexible payload. 2745 2746The Flow Director can also mask out parts of all of these fields so that filters 2747are only applied to certain fields or parts of the fields. 2748 2749Note that for raw flow type mode the source and destination fields in the 2750raw packet buffer need to be presented in a reversed order with respect 2751to the expected received packets. 2752For example: IP source and destination addresses or TCP/UDP/SCTP 2753source and destination ports 2754 2755Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2756 2757# Commands to add flow director filters of different flow types:: 2758 2759 flow_director_filter (port_id) mode IP (add|del|update) \ 2760 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2761 src (src_ip_address) dst (dst_ip_address) \ 2762 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2763 vlan (vlan_value) flexbytes (flexbytes_value) \ 2764 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2765 fd_id (fd_id_value) 2766 2767 flow_director_filter (port_id) mode IP (add|del|update) \ 2768 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2769 src (src_ip_address) (src_port) \ 2770 dst (dst_ip_address) (dst_port) \ 2771 tos (tos_value) ttl (ttl_value) \ 2772 vlan (vlan_value) flexbytes (flexbytes_value) \ 2773 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2774 fd_id (fd_id_value) 2775 2776 flow_director_filter (port_id) mode IP (add|del|update) \ 2777 flow (ipv4-sctp|ipv6-sctp) \ 2778 src (src_ip_address) (src_port) \ 2779 dst (dst_ip_address) (dst_port) \ 2780 tos (tos_value) ttl (ttl_value) \ 2781 tag (verification_tag) vlan (vlan_value) \ 2782 flexbytes (flexbytes_value) (drop|fwd) \ 2783 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2784 2785 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2786 ether (ethertype) flexbytes (flexbytes_value) \ 2787 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2788 fd_id (fd_id_value) 2789 2790 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2791 mac (mac_address) vlan (vlan_value) \ 2792 flexbytes (flexbytes_value) (drop|fwd) \ 2793 queue (queue_id) fd_id (fd_id_value) 2794 2795 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2796 mac (mac_address) vlan (vlan_value) \ 2797 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2798 flexbytes (flexbytes_value) (drop|fwd) \ 2799 queue (queue_id) fd_id (fd_id_value) 2800 2801 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2802 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2803 packet (packet file name) 2804 2805For example, to add an ipv4-udp flow type filter:: 2806 2807 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2808 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2809 fwd pf queue 1 fd_id 1 2810 2811For example, add an ipv4-other flow type filter:: 2812 2813 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2814 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2815 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2816 2817flush_flow_director 2818~~~~~~~~~~~~~~~~~~~ 2819 2820Flush all flow director filters on a device:: 2821 2822 testpmd> flush_flow_director (port_id) 2823 2824Example, to flush all flow director filter on port 0:: 2825 2826 testpmd> flush_flow_director 0 2827 2828flow_director_mask 2829~~~~~~~~~~~~~~~~~~ 2830 2831Set flow director's input masks:: 2832 2833 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2834 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2835 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2836 2837 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2838 2839 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2840 mac (mac_value) tunnel-type (tunnel_type_value) \ 2841 tunnel-id (tunnel_id_value) 2842 2843Example, to set flow director mask on port 0:: 2844 2845 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2846 src_mask 255.255.255.255 \ 2847 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2848 dst_mask 255.255.255.255 \ 2849 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2850 2851flow_director_flex_mask 2852~~~~~~~~~~~~~~~~~~~~~~~ 2853 2854set masks of flow director's flexible payload based on certain flow type:: 2855 2856 testpmd> flow_director_flex_mask (port_id) \ 2857 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2858 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2859 l2_payload|all) (mask) 2860 2861Example, to set flow director's flex mask for all flow type on port 0:: 2862 2863 testpmd> flow_director_flex_mask 0 flow all \ 2864 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2865 2866 2867flow_director_flex_payload 2868~~~~~~~~~~~~~~~~~~~~~~~~~~ 2869 2870Configure flexible payload selection:: 2871 2872 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2873 2874For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2875 2876 testpmd> flow_director_flex_payload 0 l4 \ 2877 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2878 2879get_sym_hash_ena_per_port 2880~~~~~~~~~~~~~~~~~~~~~~~~~ 2881 2882Get symmetric hash enable configuration per port:: 2883 2884 get_sym_hash_ena_per_port (port_id) 2885 2886For example, to get symmetric hash enable configuration of port 1:: 2887 2888 testpmd> get_sym_hash_ena_per_port 1 2889 2890set_sym_hash_ena_per_port 2891~~~~~~~~~~~~~~~~~~~~~~~~~ 2892 2893Set symmetric hash enable configuration per port to enable or disable:: 2894 2895 set_sym_hash_ena_per_port (port_id) (enable|disable) 2896 2897For example, to set symmetric hash enable configuration of port 1 to enable:: 2898 2899 testpmd> set_sym_hash_ena_per_port 1 enable 2900 2901get_hash_global_config 2902~~~~~~~~~~~~~~~~~~~~~~ 2903 2904Get the global configurations of hash filters:: 2905 2906 get_hash_global_config (port_id) 2907 2908For example, to get the global configurations of hash filters of port 1:: 2909 2910 testpmd> get_hash_global_config 1 2911 2912set_hash_global_config 2913~~~~~~~~~~~~~~~~~~~~~~ 2914 2915Set the global configurations of hash filters:: 2916 2917 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 2918 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 2919 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 2920 (enable|disable) 2921 2922For example, to enable simple_xor for flow type of ipv6 on port 2:: 2923 2924 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 2925 2926set_hash_input_set 2927~~~~~~~~~~~~~~~~~~ 2928 2929Set the input set for hash:: 2930 2931 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2932 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2933 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2934 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 2935 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 2936 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 2937 fld-8th|none) (select|add) 2938 2939For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 2940 2941 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 2942 2943set_fdir_input_set 2944~~~~~~~~~~~~~~~~~~ 2945 2946The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 2947on per flow type and the flexible payload. This command can be used to change input set for each flow type. 2948 2949Set the input set for flow director:: 2950 2951 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2952 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 2953 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 2954 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 2955 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 2956 sctp-dst-port|sctp-veri-tag|none) (select|add) 2957 2958For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 2959 2960 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 2961 2962global_config 2963~~~~~~~~~~~~~ 2964 2965Set different GRE key length for input set:: 2966 2967 global_config (port_id) gre-key-len (number in bytes) 2968 2969For example to set GRE key length for input set to 4 bytes on port 0:: 2970 2971 testpmd> global_config 0 gre-key-len 4 2972 2973 2974.. _testpmd_rte_flow: 2975 2976Flow rules management 2977--------------------- 2978 2979Control of the generic flow API (*rte_flow*) is fully exposed through the 2980``flow`` command (validation, creation, destruction, queries and operation 2981modes). 2982 2983Considering *rte_flow* overlaps with all `Filter Functions`_, using both 2984features simultaneously may cause undefined side-effects and is therefore 2985not recommended. 2986 2987``flow`` syntax 2988~~~~~~~~~~~~~~~ 2989 2990Because the ``flow`` command uses dynamic tokens to handle the large number 2991of possible flow rules combinations, its behavior differs slightly from 2992other commands, in particular: 2993 2994- Pressing *?* or the *<tab>* key displays contextual help for the current 2995 token, not that of the entire command. 2996 2997- Optional and repeated parameters are supported (provided they are listed 2998 in the contextual help). 2999 3000The first parameter stands for the operation mode. Possible operations and 3001their general syntax are described below. They are covered in detail in the 3002following sections. 3003 3004- Check whether a flow rule can be created:: 3005 3006 flow validate {port_id} 3007 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3008 pattern {item} [/ {item} [...]] / end 3009 actions {action} [/ {action} [...]] / end 3010 3011- Create a flow rule:: 3012 3013 flow create {port_id} 3014 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3015 pattern {item} [/ {item} [...]] / end 3016 actions {action} [/ {action} [...]] / end 3017 3018- Destroy specific flow rules:: 3019 3020 flow destroy {port_id} rule {rule_id} [...] 3021 3022- Destroy all flow rules:: 3023 3024 flow flush {port_id} 3025 3026- Query an existing flow rule:: 3027 3028 flow query {port_id} {rule_id} {action} 3029 3030- List existing flow rules sorted by priority, filtered by group 3031 identifiers:: 3032 3033 flow list {port_id} [group {group_id}] [...] 3034 3035- Restrict ingress traffic to the defined flow rules:: 3036 3037 flow isolate {port_id} {boolean} 3038 3039Validating flow rules 3040~~~~~~~~~~~~~~~~~~~~~ 3041 3042``flow validate`` reports whether a flow rule would be accepted by the 3043underlying device in its current state but stops short of creating it. It is 3044bound to ``rte_flow_validate()``:: 3045 3046 flow validate {port_id} 3047 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3048 pattern {item} [/ {item} [...]] / end 3049 actions {action} [/ {action} [...]] / end 3050 3051If successful, it will show:: 3052 3053 Flow rule validated 3054 3055Otherwise it will show an error message of the form:: 3056 3057 Caught error type [...] ([...]): [...] 3058 3059This command uses the same parameters as ``flow create``, their format is 3060described in `Creating flow rules`_. 3061 3062Check whether redirecting any Ethernet packet received on port 0 to RX queue 3063index 6 is supported:: 3064 3065 testpmd> flow validate 0 ingress pattern eth / end 3066 actions queue index 6 / end 3067 Flow rule validated 3068 testpmd> 3069 3070Port 0 does not support TCPv6 rules:: 3071 3072 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3073 actions drop / end 3074 Caught error type 9 (specific pattern item): Invalid argument 3075 testpmd> 3076 3077Creating flow rules 3078~~~~~~~~~~~~~~~~~~~ 3079 3080``flow create`` validates and creates the specified flow rule. It is bound 3081to ``rte_flow_create()``:: 3082 3083 flow create {port_id} 3084 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3085 pattern {item} [/ {item} [...]] / end 3086 actions {action} [/ {action} [...]] / end 3087 3088If successful, it will return a flow rule ID usable with other commands:: 3089 3090 Flow rule #[...] created 3091 3092Otherwise it will show an error message of the form:: 3093 3094 Caught error type [...] ([...]): [...] 3095 3096Parameters describe in the following order: 3097 3098- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3099- A matching pattern, starting with the *pattern* token and terminated by an 3100 *end* pattern item. 3101- Actions, starting with the *actions* token and terminated by an *end* 3102 action. 3103 3104These translate directly to *rte_flow* objects provided as-is to the 3105underlying functions. 3106 3107The shortest valid definition only comprises mandatory tokens:: 3108 3109 testpmd> flow create 0 pattern end actions end 3110 3111Note that PMDs may refuse rules that essentially do nothing such as this 3112one. 3113 3114**All unspecified object values are automatically initialized to 0.** 3115 3116Attributes 3117^^^^^^^^^^ 3118 3119These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3120specified before the ``pattern`` token. 3121 3122- ``group {group id}``: priority group. 3123- ``priority {level}``: priority level within group. 3124- ``ingress``: rule applies to ingress traffic. 3125- ``egress``: rule applies to egress traffic. 3126- ``transfer``: apply rule directly to endpoints found in pattern. 3127 3128Each instance of an attribute specified several times overrides the previous 3129value as shown below (group 4 is used):: 3130 3131 testpmd> flow create 0 group 42 group 24 group 4 [...] 3132 3133Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3134 3135While not specifying a direction is an error, some rules may allow both 3136simultaneously. 3137 3138Most rules affect RX therefore contain the ``ingress`` token:: 3139 3140 testpmd> flow create 0 ingress pattern [...] 3141 3142Matching pattern 3143^^^^^^^^^^^^^^^^ 3144 3145A matching pattern starts after the ``pattern`` token. It is made of pattern 3146items and is terminated by a mandatory ``end`` item. 3147 3148Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3149rte_flow_item_type``). 3150 3151The ``/`` token is used as a separator between pattern items as shown 3152below:: 3153 3154 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3155 3156Note that protocol items like these must be stacked from lowest to highest 3157layer to make sense. For instance, the following rule is either invalid or 3158unlikely to match any packet:: 3159 3160 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3161 3162More information on these restrictions can be found in the *rte_flow* 3163documentation. 3164 3165Several items support additional specification structures, for example 3166``ipv4`` allows specifying source and destination addresses as follows:: 3167 3168 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3169 dst is 10.2.0.0 / end [...] 3170 3171This rule matches all IPv4 traffic with the specified properties. 3172 3173In this example, ``src`` and ``dst`` are field names of the underlying 3174``struct rte_flow_item_ipv4`` object. All item properties can be specified 3175in a similar fashion. 3176 3177The ``is`` token means that the subsequent value must be matched exactly, 3178and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3179accordingly. Possible assignment tokens are: 3180 3181- ``is``: match value perfectly (with full bit-mask). 3182- ``spec``: match value according to configured bit-mask. 3183- ``last``: specify upper bound to establish a range. 3184- ``mask``: specify bit-mask with relevant bits set to one. 3185- ``prefix``: generate bit-mask from a prefix length. 3186 3187These yield identical results:: 3188 3189 ipv4 src is 10.1.1.1 3190 3191:: 3192 3193 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3194 3195:: 3196 3197 ipv4 src spec 10.1.1.1 src prefix 32 3198 3199:: 3200 3201 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3202 3203:: 3204 3205 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3206 3207Inclusive ranges can be defined with ``last``:: 3208 3209 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3210 3211Note that ``mask`` affects both ``spec`` and ``last``:: 3212 3213 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3214 # matches 10.1.0.0 to 10.2.255.255 3215 3216Properties can be modified multiple times:: 3217 3218 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3219 3220:: 3221 3222 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3223 3224Pattern items 3225^^^^^^^^^^^^^ 3226 3227This section lists supported pattern items and their attributes, if any. 3228 3229- ``end``: end list of pattern items. 3230 3231- ``void``: no-op pattern item. 3232 3233- ``invert``: perform actions when pattern does not match. 3234 3235- ``any``: match any protocol for the current layer. 3236 3237 - ``num {unsigned}``: number of layers covered. 3238 3239- ``pf``: match traffic from/to the physical function. 3240 3241- ``vf``: match traffic from/to a virtual function ID. 3242 3243 - ``id {unsigned}``: VF ID. 3244 3245- ``phy_port``: match traffic from/to a specific physical port. 3246 3247 - ``index {unsigned}``: physical port index. 3248 3249- ``port_id``: match traffic from/to a given DPDK port ID. 3250 3251 - ``id {unsigned}``: DPDK port ID. 3252 3253- ``mark``: match value set in previously matched flow rule using the mark action. 3254 3255 - ``id {unsigned}``: arbitrary integer value. 3256 3257- ``raw``: match an arbitrary byte string. 3258 3259 - ``relative {boolean}``: look for pattern after the previous item. 3260 - ``search {boolean}``: search pattern from offset (see also limit). 3261 - ``offset {integer}``: absolute or relative offset for pattern. 3262 - ``limit {unsigned}``: search area limit for start of pattern. 3263 - ``pattern {string}``: byte string to look for. 3264 3265- ``eth``: match Ethernet header. 3266 3267 - ``dst {MAC-48}``: destination MAC. 3268 - ``src {MAC-48}``: source MAC. 3269 - ``type {unsigned}``: EtherType or TPID. 3270 3271- ``vlan``: match 802.1Q/ad VLAN tag. 3272 3273 - ``tci {unsigned}``: tag control information. 3274 - ``pcp {unsigned}``: priority code point. 3275 - ``dei {unsigned}``: drop eligible indicator. 3276 - ``vid {unsigned}``: VLAN identifier. 3277 - ``inner_type {unsigned}``: inner EtherType or TPID. 3278 3279- ``ipv4``: match IPv4 header. 3280 3281 - ``tos {unsigned}``: type of service. 3282 - ``ttl {unsigned}``: time to live. 3283 - ``proto {unsigned}``: next protocol ID. 3284 - ``src {ipv4 address}``: source address. 3285 - ``dst {ipv4 address}``: destination address. 3286 3287- ``ipv6``: match IPv6 header. 3288 3289 - ``tc {unsigned}``: traffic class. 3290 - ``flow {unsigned}``: flow label. 3291 - ``proto {unsigned}``: protocol (next header). 3292 - ``hop {unsigned}``: hop limit. 3293 - ``src {ipv6 address}``: source address. 3294 - ``dst {ipv6 address}``: destination address. 3295 3296- ``icmp``: match ICMP header. 3297 3298 - ``type {unsigned}``: ICMP packet type. 3299 - ``code {unsigned}``: ICMP packet code. 3300 3301- ``udp``: match UDP header. 3302 3303 - ``src {unsigned}``: UDP source port. 3304 - ``dst {unsigned}``: UDP destination port. 3305 3306- ``tcp``: match TCP header. 3307 3308 - ``src {unsigned}``: TCP source port. 3309 - ``dst {unsigned}``: TCP destination port. 3310 3311- ``sctp``: match SCTP header. 3312 3313 - ``src {unsigned}``: SCTP source port. 3314 - ``dst {unsigned}``: SCTP destination port. 3315 - ``tag {unsigned}``: validation tag. 3316 - ``cksum {unsigned}``: checksum. 3317 3318- ``vxlan``: match VXLAN header. 3319 3320 - ``vni {unsigned}``: VXLAN identifier. 3321 3322- ``e_tag``: match IEEE 802.1BR E-Tag header. 3323 3324 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3325 3326- ``nvgre``: match NVGRE header. 3327 3328 - ``tni {unsigned}``: virtual subnet ID. 3329 3330- ``mpls``: match MPLS header. 3331 3332 - ``label {unsigned}``: MPLS label. 3333 3334- ``gre``: match GRE header. 3335 3336 - ``protocol {unsigned}``: protocol type. 3337 3338- ``fuzzy``: fuzzy pattern match, expect faster than default. 3339 3340 - ``thresh {unsigned}``: accuracy threshold. 3341 3342- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3343 3344 - ``teid {unsigned}``: tunnel endpoint identifier. 3345 3346- ``geneve``: match GENEVE header. 3347 3348 - ``vni {unsigned}``: virtual network identifier. 3349 - ``protocol {unsigned}``: protocol type. 3350 3351- ``vxlan-gpe``: match VXLAN-GPE header. 3352 3353 - ``vni {unsigned}``: VXLAN-GPE identifier. 3354 3355- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3356 3357 - ``sha {MAC-48}``: sender hardware address. 3358 - ``spa {ipv4 address}``: sender IPv4 address. 3359 - ``tha {MAC-48}``: target hardware address. 3360 - ``tpa {ipv4 address}``: target IPv4 address. 3361 3362- ``ipv6_ext``: match presence of any IPv6 extension header. 3363 3364 - ``next_hdr {unsigned}``: next header. 3365 3366- ``icmp6``: match any ICMPv6 header. 3367 3368 - ``type {unsigned}``: ICMPv6 type. 3369 - ``code {unsigned}``: ICMPv6 code. 3370 3371- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3372 3373 - ``target_addr {ipv6 address}``: target address. 3374 3375- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3376 3377 - ``target_addr {ipv6 address}``: target address. 3378 3379- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3380 3381 - ``type {unsigned}``: ND option type. 3382 3383- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3384 link-layer address option. 3385 3386 - ``sla {MAC-48}``: source Ethernet LLA. 3387 3388- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3389 link-layer address option. 3390 3391 - ``tla {MAC-48}``: target Ethernet LLA. 3392 3393Actions list 3394^^^^^^^^^^^^ 3395 3396A list of actions starts after the ``actions`` token in the same fashion as 3397`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3398terminated by a mandatory ``end`` action. 3399 3400Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3401rte_flow_action_type``). 3402 3403Dropping all incoming UDPv4 packets can be expressed as follows:: 3404 3405 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3406 actions drop / end 3407 3408Several actions have configurable properties which must be specified when 3409there is no valid default value. For example, ``queue`` requires a target 3410queue index. 3411 3412This rule redirects incoming UDPv4 traffic to queue index 6:: 3413 3414 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3415 actions queue index 6 / end 3416 3417While this one could be rejected by PMDs (unspecified queue index):: 3418 3419 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3420 actions queue / end 3421 3422As defined by *rte_flow*, the list is not ordered, all actions of a given 3423rule are performed simultaneously. These are equivalent:: 3424 3425 queue index 6 / void / mark id 42 / end 3426 3427:: 3428 3429 void / mark id 42 / queue index 6 / end 3430 3431All actions in a list should have different types, otherwise only the last 3432action of a given type is taken into account:: 3433 3434 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3435 3436:: 3437 3438 drop / drop / drop / end # drop is performed only once 3439 3440:: 3441 3442 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3443 3444Considering they are performed simultaneously, opposite and overlapping 3445actions can sometimes be combined when the end result is unambiguous:: 3446 3447 drop / queue index 6 / end # drop has no effect 3448 3449:: 3450 3451 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3452 3453:: 3454 3455 drop / passthru / end # drop has no effect 3456 3457Note that PMDs may still refuse such combinations. 3458 3459Actions 3460^^^^^^^ 3461 3462This section lists supported actions and their attributes, if any. 3463 3464- ``end``: end list of actions. 3465 3466- ``void``: no-op action. 3467 3468- ``passthru``: let subsequent rule process matched packets. 3469 3470- ``jump``: redirect traffic to group on device. 3471 3472 - ``group {unsigned}``: group to redirect to. 3473 3474- ``mark``: attach 32 bit value to packets. 3475 3476 - ``id {unsigned}``: 32 bit value to return with packets. 3477 3478- ``flag``: flag packets. 3479 3480- ``queue``: assign packets to a given queue index. 3481 3482 - ``index {unsigned}``: queue index to use. 3483 3484- ``drop``: drop packets (note: passthru has priority). 3485 3486- ``count``: enable counters for this rule. 3487 3488- ``rss``: spread packets among several queues. 3489 3490 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3491 the same as `set_hash_global_config`_. 3492 3493 - ``level {unsigned}``: encapsulation level for ``types``. 3494 3495 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3496 tokens are the same as `set_hash_input_set`_, except that an empty list 3497 does not disable RSS but instead requests unspecified "best-effort" 3498 settings. 3499 3500 - ``key {string}``: RSS hash key, overrides ``key_len``. 3501 3502 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3503 conjunction with ``key`` to pad or truncate it. 3504 3505 - ``queues [{unsigned} [...]] end``: queue indices to use. 3506 3507- ``pf``: direct traffic to physical function. 3508 3509- ``vf``: direct traffic to a virtual function ID. 3510 3511 - ``original {boolean}``: use original VF ID if possible. 3512 - ``id {unsigned}``: VF ID. 3513 3514- ``phy_port``: direct packets to physical port index. 3515 3516 - ``original {boolean}``: use original port index if possible. 3517 - ``index {unsigned}``: physical port index. 3518 3519- ``port_id``: direct matching traffic to a given DPDK port ID. 3520 3521 - ``original {boolean}``: use original DPDK port ID if possible. 3522 - ``id {unsigned}``: DPDK port ID. 3523 3524- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3525 3526 - ``mpls_ttl``: MPLS TTL. 3527 3528- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3529 3530- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3531 3532 - ``nw_ttl``: IP TTL. 3533 3534- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3535 3536- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3537 3538- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3539 3540- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3541 3542- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3543 3544 - ``ethertype``: Ethertype. 3545 3546- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3547 3548 - ``vlan_vid``: VLAN id. 3549 3550- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3551 3552 - ``vlan_pcp``: VLAN priority. 3553 3554- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3555 3556 - ``ethertype``: Ethertype. 3557 3558- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3559 3560 - ``ethertype``: Ethertype. 3561 3562Destroying flow rules 3563~~~~~~~~~~~~~~~~~~~~~ 3564 3565``flow destroy`` destroys one or more rules from their rule ID (as returned 3566by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3567times as necessary:: 3568 3569 flow destroy {port_id} rule {rule_id} [...] 3570 3571If successful, it will show:: 3572 3573 Flow rule #[...] destroyed 3574 3575It does not report anything for rule IDs that do not exist. The usual error 3576message is shown when a rule cannot be destroyed:: 3577 3578 Caught error type [...] ([...]): [...] 3579 3580``flow flush`` destroys all rules on a device and does not take extra 3581arguments. It is bound to ``rte_flow_flush()``:: 3582 3583 flow flush {port_id} 3584 3585Any errors are reported as above. 3586 3587Creating several rules and destroying them:: 3588 3589 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3590 actions queue index 2 / end 3591 Flow rule #0 created 3592 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3593 actions queue index 3 / end 3594 Flow rule #1 created 3595 testpmd> flow destroy 0 rule 0 rule 1 3596 Flow rule #1 destroyed 3597 Flow rule #0 destroyed 3598 testpmd> 3599 3600The same result can be achieved using ``flow flush``:: 3601 3602 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3603 actions queue index 2 / end 3604 Flow rule #0 created 3605 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3606 actions queue index 3 / end 3607 Flow rule #1 created 3608 testpmd> flow flush 0 3609 testpmd> 3610 3611Non-existent rule IDs are ignored:: 3612 3613 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3614 actions queue index 2 / end 3615 Flow rule #0 created 3616 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3617 actions queue index 3 / end 3618 Flow rule #1 created 3619 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3620 testpmd> 3621 testpmd> flow destroy 0 rule 0 3622 Flow rule #0 destroyed 3623 testpmd> 3624 3625Querying flow rules 3626~~~~~~~~~~~~~~~~~~~ 3627 3628``flow query`` queries a specific action of a flow rule having that 3629ability. Such actions collect information that can be reported using this 3630command. It is bound to ``rte_flow_query()``:: 3631 3632 flow query {port_id} {rule_id} {action} 3633 3634If successful, it will display either the retrieved data for known actions 3635or the following message:: 3636 3637 Cannot display result for action type [...] ([...]) 3638 3639Otherwise, it will complain either that the rule does not exist or that some 3640error occurred:: 3641 3642 Flow rule #[...] not found 3643 3644:: 3645 3646 Caught error type [...] ([...]): [...] 3647 3648Currently only the ``count`` action is supported. This action reports the 3649number of packets that hit the flow rule and the total number of bytes. Its 3650output has the following format:: 3651 3652 count: 3653 hits_set: [...] # whether "hits" contains a valid value 3654 bytes_set: [...] # whether "bytes" contains a valid value 3655 hits: [...] # number of packets 3656 bytes: [...] # number of bytes 3657 3658Querying counters for TCPv6 packets redirected to queue 6:: 3659 3660 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3661 actions queue index 6 / count / end 3662 Flow rule #4 created 3663 testpmd> flow query 0 4 count 3664 count: 3665 hits_set: 1 3666 bytes_set: 0 3667 hits: 386446 3668 bytes: 0 3669 testpmd> 3670 3671Listing flow rules 3672~~~~~~~~~~~~~~~~~~ 3673 3674``flow list`` lists existing flow rules sorted by priority and optionally 3675filtered by group identifiers:: 3676 3677 flow list {port_id} [group {group_id}] [...] 3678 3679This command only fails with the following message if the device does not 3680exist:: 3681 3682 Invalid port [...] 3683 3684Output consists of a header line followed by a short description of each 3685flow rule, one per line. There is no output at all when no flow rules are 3686configured on the device:: 3687 3688 ID Group Prio Attr Rule 3689 [...] [...] [...] [...] [...] 3690 3691``Attr`` column flags: 3692 3693- ``i`` for ``ingress``. 3694- ``e`` for ``egress``. 3695 3696Creating several flow rules and listing them:: 3697 3698 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3699 actions queue index 6 / end 3700 Flow rule #0 created 3701 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3702 actions queue index 2 / end 3703 Flow rule #1 created 3704 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3705 actions rss queues 6 7 8 end / end 3706 Flow rule #2 created 3707 testpmd> flow list 0 3708 ID Group Prio Attr Rule 3709 0 0 0 i- ETH IPV4 => QUEUE 3710 1 0 0 i- ETH IPV6 => QUEUE 3711 2 0 5 i- ETH IPV4 UDP => RSS 3712 testpmd> 3713 3714Rules are sorted by priority (i.e. group ID first, then priority level):: 3715 3716 testpmd> flow list 1 3717 ID Group Prio Attr Rule 3718 0 0 0 i- ETH => COUNT 3719 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3720 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3721 1 24 0 i- ETH IPV4 UDP => QUEUE 3722 4 24 10 i- ETH IPV4 TCP => DROP 3723 3 24 20 i- ETH IPV4 => DROP 3724 2 24 42 i- ETH IPV4 UDP => QUEUE 3725 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3726 testpmd> 3727 3728Output can be limited to specific groups:: 3729 3730 testpmd> flow list 1 group 0 group 63 3731 ID Group Prio Attr Rule 3732 0 0 0 i- ETH => COUNT 3733 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3734 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3735 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3736 testpmd> 3737 3738Toggling isolated mode 3739~~~~~~~~~~~~~~~~~~~~~~ 3740 3741``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3742must only be injected from the defined flow rules; that no default traffic 3743is expected outside those rules and the driver is free to assign more 3744resources to handle them. It is bound to ``rte_flow_isolate()``:: 3745 3746 flow isolate {port_id} {boolean} 3747 3748If successful, enabling or disabling isolated mode shows either:: 3749 3750 Ingress traffic on port [...] 3751 is now restricted to the defined flow rules 3752 3753Or:: 3754 3755 Ingress traffic on port [...] 3756 is not restricted anymore to the defined flow rules 3757 3758Otherwise, in case of error:: 3759 3760 Caught error type [...] ([...]): [...] 3761 3762Mainly due to its side effects, PMDs supporting this mode may not have the 3763ability to toggle it more than once without reinitializing affected ports 3764first (e.g. by exiting testpmd). 3765 3766Enabling isolated mode:: 3767 3768 testpmd> flow isolate 0 true 3769 Ingress traffic on port 0 is now restricted to the defined flow rules 3770 testpmd> 3771 3772Disabling isolated mode:: 3773 3774 testpmd> flow isolate 0 false 3775 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3776 testpmd> 3777 3778Sample QinQ flow rules 3779~~~~~~~~~~~~~~~~~~~~~~ 3780 3781Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3782 3783 testpmd> port stop 0 3784 testpmd> vlan set qinq on 0 3785 3786The above command sets the inner and outer TPID's to 0x8100. 3787 3788To change the TPID's the following commands should be used:: 3789 3790 testpmd> vlan set outer tpid 0xa100 0 3791 testpmd> vlan set inner tpid 0x9100 0 3792 testpmd> port start 0 3793 3794Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3795 3796:: 3797 3798 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3799 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3800 Flow rule #0 validated 3801 3802 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3803 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3804 Flow rule #0 created 3805 3806 testpmd> flow list 0 3807 ID Group Prio Attr Rule 3808 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3809 3810Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3811 3812:: 3813 3814 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3815 vlan tci is 654 / end actions pf / queue index 0 / end 3816 Flow rule #1 validated 3817 3818 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3819 vlan tci is 654 / end actions pf / queue index 1 / end 3820 Flow rule #1 created 3821 3822 testpmd> flow list 0 3823 ID Group Prio Attr Rule 3824 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3825 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3826